Levine on Wall Street: The Mystery of Mortgage Fraud

Reversing the usual trend, low-level mortgage fraudsters get off, while Goldman pays billions in settlements. Also: Valeant/Allergan, CDS terms, Goldman partners, and the Harvard Club.

What is mortgage fraud, anyway?

Here's how the FBI puts it (not legal advice?): It's "some type of material misstatement, misrepresentation, or omission relating to a real estate transaction which is relied on by one or more parties to the transaction." One thing that looks like it might be mortgage fraud is: X borrows $100 to buy a $100 house, misrepresenting his income and intent to live in that house. X then sells the house to Y at an inflated price, say $150, and Y borrows $150 to pay the purchase price by similarly misrepresenting his income, intentions, etc. X has now made $50, which he splits with Y, and then they default on both mortgages and walk away with their $50. A Sacramento real estate agent named Yevgenity Charikov and his friends were accused of running more or less this scam in 2006 and 2007.

But hang on, Charikov said: Actually banks in 2006-2007 didn't care at all about anyone's income, or their intent to live in any houses. They were just churning out mortgages for the securitization market. They'd make loans to anyone. Sure, maybe our mortgage applications were full of lies, but those lies weren't material, and no one relied on them. So we couldn't have committed criminal mortgage fraud.

This argument worked: On Friday a federal jury acquitted Charikov and his friends. It also strikes me as obviously correct, and sort of wonderful, and something that everyone should be happy about. Charikov's lawyers took a populist view, arguing to the jury and anyone else who would listen that the real culprits are the evil banksters etc. etc. But if you're an evil bankster you should like this result even more. After all, when the evil banksters packaged and sold these mortgages to investors, those packages were also full of lies. And so those banks are out tens of billions of dollars for fraud settlements. But hang on, they might say: Sure we lied a lot about those mortgages. But where is the evidence that anyone cared? The mortgage machine needed to be fed, and we fed it, just like this Charikov character. It's only fraud if the victim wasn't in on it.

And what does it cost?

Elsewhere in mortgage fraud, Goldman agreed to pay $3.15 billion to buy back mortgage-backed securities worth $1.95 billion from Fannie Mae and Freddie Mac, to settle Federal Housing Finance Agency fraud claims. So that's $1.2 billion if you're keeping score I guess. The Financial Times is keeping score, and wonders how the Department of Justice, other regulators (FHFA, state attorneys general, etc.), and the banks figure out the numbers that go on these cases. "It is a black box," says a bank lawyer, and that seems about right. I complained last week about the sheer repetitiveness of these settlements -- Bank of America has settled its Fannie Mae claims five times now -- and the opacity is similarly frustrating. There's no clarity of "here's what they did, here's how much of it they did, and here's how much it costs them." It's just "here's what we were able to negotiate this time, isn't it big?"

And how do we get rid of it?

The Securities and Exchange Commission wants to require more disclosure about borrowers' income and credit scores in asset-backed securities. But here's a good post from Adam Levitin about subprime auto lending and ability-to-pay requirements. His argument is that "The advent of securitization and the sweatbox business model have broken the lender-borrower partnership." Securitization is what we were just talking about: If you're lending money to people for like a minute before you securitize the loans and sell them to muppets, you don't actually care if the people will pay you back, so you can just accept any old lies about their income. This destroys the old notion -- "the lender-borrower partnership" -- that a bank has good incentives only to lend money to people who can pay it back. It's no longer the bank's money. And then the "sweatbox model involves a lender that is willing to treat the loan's principal as a loss leader: if the lender collects enough in interest and fees over time, the lender may still make a profit even with the loss of some of the principal." Levitin's proposed fix is to forbid banks from making loans unless they think that they'll get paid back, which sounds super obvious and unnecessary until you look at how loans are actually made.

Allergan will hold a special meeting to vote on the Valeant deal (probably).

Bill Ackman delivered 1,500 pages of documents to Allergan, representing requests from holders of more than 30 percent of Allergan's shares to hold a special meeting of shareholders to vote on the takeover proposal from Valeant Pharmaceuticals and Pershing Square. Here is Pershing Square's press release, which says that "Allergan is now required to call the meeting no later than December 20, 2014," though in the next sentence it adds that Pershing Square and Valeant are preemptively suing to make it do that. Here is Allergan's press release, which puts a brave face on things ("many stockholders have explicitly conveyed their view that the requests are not an endorsement of Valeant's offer"), and promises to review the requests and get back to everyone. In 1,500 pages of forms you can probably find something filled out wrong, and the question is, if there is a mistake, will Allergan be petty enough to deny shareholders their special meeting? Given the nastiness of this battle so far, I can't wait to find out the answer.

There are new ISDA Credit Definitions.

Are you not excited? Sure you are, come on. Awesome new provisions for the way we credit default swap now include:

  • "a new Credit Event, Government Intervention, with respect to non-U.S. financial Reference Entities," so when a European government does some weird thing to a bank but doesn't call it default, you can still trigger CDS; and
  • "Asset Package Delivery," so when your bank or sovereign poofs its bonds into some package of other weird things -- new bonds, say, or equity, or amorphous "claims" -- you can just hand that package of weird things in to the CDS auction.

There won't be that many Goldman Sachs partners this year.

This year's partner class will be "no more than 70 employees," "people familiar with the matter" told the Wall Street Journal. It would be funny if those people familiar with the matter also dropped a few names, like, "no more than 70 people, Jane Smith is looking pretty likely, but John Doe has no shot."

Things happen.

Companies are buying back less stock as prices go up. If a Dutch company buys an American one it doesn't have to pay American corporate taxes any more, even if the Dutch company is basically Bain Capital. Burger King might buy Tim Hortons. "The issue of cyber resilience is a bit of a sleeper issue." A Leverage-Based Measure of Financial Instability. Investment-grade whisky. It's getting easier to get drunk at the Harvard Club.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

    To contact the author on this story:
    Matthew S Levine at mlevine51@bloomberg.net

    To contact the editor on this story:
    Toby Harshaw at tharshaw@bloomberg.net

    Before it's here, it's on the Bloomberg Terminal.